As if on cue ...

[hrj]

Thursday I was at this all-day training session at work for "managing multiple priorities". Among the other dumb and questionable things the lecturer said was, "a scientific study has proven that e-mail and texting interruptions lower the IQ twice as much as smoking marijuana." My instant reaction was, "This is just like those pseudo-scientific sound-bite studies that the folks at Language Log are always frothing at the mouth about." (I.e., sociological or behavioral studies with extremely marginal distinctions between categories of individuals based on small data sets where not only is the actual difference statistically insignificant, but is much smaller than each category's internal variation.) I want to see data. I want to see controls. I want to see the study design. I want to see statistical significance.

And then, lo and behold, today's Language Log column makes reference to a column they'd done debunking this exact study all the way back in 2005. Hah! Question #1: Is it worth my time to direct the attention of the trainer to this debunking? Question #2: If I do, is it likely to result in any change in his obviously canned patter?

Comments

[ermine-rat.livejournal.com]

You could tell him that 90% of statistics are made up on the spot...

"there are lies, damn lies, and statistics."
---Twain